Guide to the TechWeb Network

Intelligent Enterprise

Better Insight for Business Decisions

Intelligent Enterprise - Better Insight for Business Decisions
search Intelligent Enterprise
Advanced Search
RSS
Webcasts
Whitepapers
Subscribe
Home




March 28, 2002

/020328/506decision1_1.jhtml">

The Semantic Web

A semantic Web will enable automated use of disparate, distributed Internet information sources and services

By Seth Grimes

Continued from Page 1

Our interests in the decision-support world will be best served by furthering distributed semantic Web development in preference to unnecessary centralization. This direction will offer greater flexibility and openness. And, it would not preclude the use of query and information-interchange protocols, such as the simple object access protocol, that form an independent part of the Web services portfolio.

Realizing the Semantic Web

The semantic Web relies on three technologies filling key roles:

  • XML for syntax and structure
  • Ontology systems that define terms and their relationships
  • The Resource Definition Framework (RDF), which provides a model for encoding ontology defined meaning.

Other technologies and concepts also come into play: universal resource identifiers, which are globally recognized and unique element definitions, rules-processing (inference) systems, and the usual Internet-infrastructure protocols.

An ontology starts with a taxonomy, a structured arrangement of information into classes that categorizes a subject area and relates its elements. Many ontology projects are underway. (If you want to check one out, try OpenCyc, which I first encountered while researching "Out in the Open", which discussed open-source decision-support tools. OpenCyc is a planned subset of Cycorp's general-knowledge ontology. Because it's open source, you can freely use and help extend it.)

The semantic Web effort mandates that you express ontologies in RDF using XML. XML forms a strong foundation for modern, layered approaches to constructing markup language, and it's great for syntax. But without RDF (or an equivalent), XML-based constructs lack meaning.

The RDF specification envisions an object-oriented system of classes forming a schema. According to the specification, three types of objects make up the basic data model:

  • Resources designated by URLs for any entity
  • Properties, each representing a specific aspect, characteristic, attribute, or relation, to describe a resource
  • Statements, each having a subject resource, an object resource or literal, and a predicate that ties subject to object.



Rate This Article

Comments:

Optional e-mail address:

The semantic Web will be realized through the continued development of standards and technologies including those I've described and others such as the DARPA agent markup language (DAML). DAML is a family of markup languages providing the power to express ontologies that use RDF to provide semantic meaning. DAML languages incorporate artificial intelligence knowledge-representation concepts and are designed to support agent and inference-engine interactions with suitable marked-up sites. Tools such as W3C's Annotea collaborative metadata-annotation system are essential in completing the picture of a web of semantically defined services.

Directions

Do you have visions of analytic agents crawling the Web to locate and negotiate access to a spectrum of up-to-the minute data sources, using inference rules to judge the value and quality of their findings, and ontologies to transform findings into a harmonized form suitable for analysis? Maybe not - there's no special reason you should think in terms of particular implementation technologies - but you probably do like the idea of automated, closed-loop decision systems that reach beyond centralized directories to the whole of the Web.

If you're a data or service provider, think about joining efforts to create ontologies and XML schemas for your subject area. Learn about DAML and how you can expand your services to support automated agents. If you write analytic software, think about extending your products to operate in the semantic Web. Although semantic Web technologies are still several years from widespread implementation, they show potential for adding enormous value to the Internet. Providers and users of analytic data and services, given their understanding of metadata, KM, and agent technology, are ideally positioned to be among the first to benefit.


Seth Grimes [grimes@altaplana.com] is a principal of Alta Plana Corp., a Washington, D.C.-based consultancy specializing in analytic computing systems and demographic and economic statistics.


RESOURCES

"Semantic Web Services": www.daml.org/services/ieee01-KSL.pdf

"Taxonomies Put Content in Context": www.transformmag.com/db_area/archs/2001/12/tfm0112f1.jhtml

"The Semantic Web": scientificamerican.com/2001/0501issue/0501berners-lee.html

"Web Search Your Own Way": Journal of the Association for Computing Machinery, December 2001

Dublin Core Metadata Initiative: www.dublincore.org

OpenCyc Ontology Project: opencyc.org

Resource Description Framework Model and Syntax Specification: www.w3.org/TR/1999/REC-rdf-syntax-19990222

The World Wide Web Consortium Semantic Web: w3.org/2001/sw








IE Weekly Newsletter
Subscribe to the newsletter
    Email Address







techweb
Online Communities TechWebInformationWeekLight ReadingIntelligent EnterprisebMightyNetwork ComputingDark ReadingDigital LibraryWall Street & Technology
Byte & SwitchNo JitterInternet EvolutionLight Reading's Cable Digital NewsContentinopleUnStrungBank Systems & TechnologyAdvanced TradingInsurance & Technology
Face-to-Face Events
InteropWeb 2.0 ExpoWeb 2.0 SummitVoiceConBlack HatCSISoftwareEntrprise 2.0 ConferenceGTEC
Mobile Business Expo
InformationWeek 500 ConferenceBuy Side Trading XchangeBuy Side Trading SummitBank Executive SummitInsurance Executive SummitTelcoTVEthernet ExpoOptical Expo
Magazines  
InformationWeekWall Street & TechnologyInsurance & TechnologyBank Systems & TechnologyAdvanced TradingMSDNTechNetSmart EnterpriseThe Architecture JournalDatabase Magazine
 
Research & Analyst Services  
Heavy ReadingInformationWeek ReportsInformationWeek Analytics